fix: delete agent 500 error + dynamic personality + deployment guide

- Fix delete agent 500: clean up FK records (agent_llm_logs, permissions,
  schedules, executions, team_members) and unbind goals/tasks before delete
- Remove hardcoded personality templates in Android, replace with dynamic
  system prompt generation from name + description
- Set promptSectionsEnabled=false to bypass PromptComposer for personality
- Add Tencent Cloud Linux deployment guide (Docker Compose)
- Accumulated backend service updates, frontend UI fixes, Android app changes

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

backend/app/agent_runtime/memory.py

@@ -44,6 +44,7 @@ class AgentMemory:
         team_share_enabled: bool = False,
         memory_dir_enabled: bool = False,
         memory_dir_path: str = "",
+        parent_agent_id: Optional[str] = None,
     ):
         self.scope_kind = scope_kind
         self.scope_id = scope_id or "default"
@@ -54,6 +55,7 @@ class AgentMemory:
         self.vector_memory_top_k = vector_memory_top_k
         self.vector_memory_rerank = vector_memory_rerank
         self.memory_type_filter = memory_type_filter  # None = 全部类型
+        self.parent_agent_id = parent_agent_id  # 父 Agent ID（继承经验）
         self.team_id = team_id  # 团队共享 ID
         self.team_share_enabled = team_share_enabled  # 是否自动发布到团队池
         # 文件式记忆
@@ -138,6 +140,12 @@ class AgentMemory:
         if global_text:
             parts.append(global_text)
 
+        # 5. 父 Agent 知识继承：加载父级经验
+        if self.parent_agent_id and self.scope_id:
+            parent_text = await self._parent_knowledge_search(query)
+            if parent_text:
+                parts.append(parent_text)
+
         return "\n\n".join(parts) if parts else ""
 
     async def _vector_search(self, query: str = "") -> str:
@@ -341,6 +349,65 @@ class AgentMemory:
             logger.warning("LLM Rerank 失败，使用向量排序: %s", e)
             return candidates[: self.vector_memory_top_k]
 
+    async def _parent_knowledge_search(self, query: str = "") -> str:
+        """从父 Agent 的知识库中检索相关经验。"""
+        from app.models.knowledge_entry import KnowledgeEntry
+        from app.models.agent import Agent
+
+        db: Optional[Session] = None
+        try:
+            db = SessionLocal()
+
+            # 向上追溯祖先 Agent（最多 3 层）
+            ancestor_ids: List[str] = []
+            current_id = self.parent_agent_id
+            for _ in range(3):
+                if not current_id:
+                    break
+                agent = db.query(Agent).filter(Agent.id == current_id).first()
+                if not agent:
+                    break
+                ancestor_ids.append(current_id)
+                current_id = getattr(agent, "parent_agent_id", None)
+                if current_id in ancestor_ids:
+                    break
+
+            if not ancestor_ids:
+                return ""
+
+            # 查询祖先的知识条目
+            entries = (
+                db.query(KnowledgeEntry)
+                .filter(
+                    KnowledgeEntry.agent_id.in_(ancestor_ids),
+                    KnowledgeEntry.is_active == True,
+                    KnowledgeEntry.confidence >= 0.5,
+                )
+                .order_by(KnowledgeEntry.retrieval_count.desc())
+                .limit(8)
+                .all()
+            )
+
+            if not entries:
+                return ""
+
+            lines = [f"## 父级 Agent 经验（来自 {len(ancestor_ids)} 个祖先 Agent）"]
+            for i, entry in enumerate(entries, 1):
+                cat = entry.category or "经验"
+                title = entry.title or ""
+                sol = (entry.solution or entry.situation or "")[:300]
+                caveat = f" 注意: {entry.caveats[:100]}" if entry.caveats else ""
+                lines.append(f"{i}. [{cat}] {title}: {sol}{caveat}")
+
+            return "\n".join(lines)
+
+        except Exception as e:
+            logger.warning("父级知识检索失败: %s", e)
+            return ""
+        finally:
+            if db:
+                db.close()
+
     async def _global_knowledge_search(self, query: str = "") -> str:
         """从 GlobalKnowledge 表检索相关的全局知识条目。"""
         from datetime import datetime